A prominent Irish Language enthusiast has gone to great pains to tell us Irish was once widely spoken in East Belfast according to Census figures. This new source of information becomes available to us with the Census of Ireland of 1851, in which a language question was included for the first time. This was largely due to the prompting of Belfastman Robert McAdam, and the question was included in every census from 1851 until 1911, that is, every ten years during that time.These figures show surprising fluctuations and have to be carefully considered, but they are of the utmost interest all the same. The problems with the language question were many. The question simply asked whether the subject spoke Irish well, or whether as one’s first language, was simply not asked. Varying social pressures also affected the responses.

This is how MacAdam describes the attitude prevailing in 1851: “In various districts where the two languages coexist, but where English now largely predominates, numbers of individuals returned themselves as ignorant of the Irish language, either from a sort of false shame, or from a secret dread that the Government in making this inquiry (for the first time) had some concealed motive, which could not be for their good. Their native shrewdness, therefore, dictated to them that their safest policy was to appear ignorant of the unfashionable language.”

The upturn in 1881 is doubtless due to the fact that the language question had been until then merely a footnote on the census form, and could easily be ignored by the enumerators. In the last two censuses we see the effects of the Gaelic revival The Gaelic League was founded in 1893, and people began in numbers to learn Irish as a second language. Adams makes a useful distinction when he divides Irish-speaking as reported in these censuses into three categories: survival Irish, the original Gaelic of the district, on which I will be focussing; immigration Irish, due to inward population movement from Gaelic-speaking districts elsewhere, as we might expect to find in urban centres; and revival Irish, learned as a second language in classes, which may be quite different from the original local form, and which dominates the 1911 figures.

The original forms for the 1901 and 1911 censuses are available for public inspection, though those for earlier censuses have been lost. When the forms for 1901 are examined, even greater doubt is cast on the meaning of the published figures and it has proved impossible to determine which individuals were included as Irish speakers in the published table. This was because the responses to the language question were in many cases amended, whether by the householders, by the enumerators, or by the census staff. There may be several layers of amendments on one form, making the response practically illegible. It can also happen that the reply is ambiguous, for example, when it is filled in for the head of household only, with dots or ticks for the other members. Even the clear responses are often absurd, such as when a family is made to consist of a number of Irish-speaking monoglots together with a number of English-speaking monoglots.

But most perplexing of all is the large number of Protestant families who claimed to be monoglot Irish speakers, from their oldest to their youngest member. Roman Catholic Gaelic-speakers, on the other hand, almost always claimed to be bilingual, and rarely included children. It seems clear that the Protestant families who returned themselves as monoglot Irish speakers misunderstood the question; they took the term ˜Irish language” to refer to Hiberno-English, or perhaps they did not see the question in linguistic terms at all. The forms have been subjected to a systematic attempt to delete those ˜Irish” and ˜Irish and English” responses which may be regarded as errors. The response should therefore have to be cancelled for Protestants, for whole families, for those below middle age, and for those claiming to be monoglots. Thus the responses to the language question, at least in Belfast in 1901, are seen to be very unreliable.